(FORTUNE Magazine) – Your correspondent recently had a choice between a weekend in which he could (a) defy the dermatologist and loll on the white beaches of North Carolina's Outer Banks or (b) defy common sense and visit highly humid Washington, D.C., where a multitude of antitest fanatics was running the shindig put on by FairTest, a.k.a. the National Center for Fair & Open Testing. Now you know where we got the tan. Luckily we have received quite detailed reports on the FairTest extravaganza. (A few moles were present.) We know, for example, that featured speaker Albert Gore, erstwhile presidential candidate, made the requisite remarks on behalf of poor youngsters who are qualified for good careers but set back in life by unfair test questions. Oddly enough, Al seemed not to have heard about any qualified poor kids whose high test scores represent their big chance for admission to good colleges. Evidencing not a whole lot of original research, he complained in particular about an SAT analogy question in which the kids are asked which of several possible pairs is comparable to ''runner: marathon.'' The correct answer is ''oarsman: regatta.'' The Senator's audience didn't need much elaboration of this particular question, since FairTest literature has complained about it more than once in recent years (even though the question has not been used since 1973). The complaint is that it's unfair, and culturally biased, because minority kids don't spend much time at upscale affairs like regattas. A detail never mentioned by FairTest is that black kids did about as well on the regatta question as did white kids with similar total scores. What does FairTest want? The answer, obviously, is affirmative action in testing. Translation: It wants test designers like the Educational Testing Service (ETS) to produce tests on which the sexes and different ethnic groups will dependably have average scores that are less far apart than at present. FairTest spokesmen tell us that the organization is not demanding ''equality of result'' in tests; however, it endlessly points to inequality of result as evidence of bias. FairTest came into being as a result of the infamous Golden Rule settlement. In 1976, ETS was sued because of a test it had designed to help select insurance agents. The test, used by the Golden Rule Insurance Co. in Illinois, had resulted in fewer minority-group members being hired as agents. In 1984, ETS decided to settle the eight-year-old case rather than fight, and lamely agreed to affirmative action -- i.e., to redesign the test in ways that would predictably limit differences in average group scores. Golden Rule, meanwhile, agreed to fund FairTest. One of the organization's major preoccupations in the years since it was formed has been to extend the Golden Rule concept. The organization claims to have ''led the national campaign to reduce the discriminatory impact of tests . . . and educate decision-makers . . . about the problems of standardized . . . exams.'' Just as you might expect, the average politician instantly understands that there are more votes in bias prevention than in sustaining the integrity of the testing process. Early last year ETS made a quite extraordinary apology to the American Psychological Association. In a letter published in an APA bulletin, ETS President Gregory R. Anrig said it had been a mistake to settle the Golden Rule case. He also said it was increasingly difficult to uphold ''sound psychometric standards'' with politicians crawling all over the tests. We may never see the regatta question again.